TiDB Cloud
PingCAP · Other
TiDB Cloud is a fully managed distributed SQL database from PingCAP. It's a MySQL-compatible database that scales without manual sharding and runs transactional and analytical queries on one cluster.
About TiDB Cloud
What Is TiDB Cloud
TiDB Cloud is PingCAP's managed version of TiDB, an open-source database that splits storage and compute so a single system handles both everyday app writes and heavy analytical queries. You create a cluster in the console, connect with the same MySQL drivers and ORMs you already use, and let the platform deal with replication, failover, and patches. There's no install and no server to babysit.
The problem it targets is the wall a normal database hits when traffic grows. Once a single MySQL instance runs out of headroom, teams usually shard by hand or bolt on a separate analytics warehouse and keep the two in sync. TiDB Cloud skips both steps. Data distributes across nodes on its own, transactions stay ACID and strongly consistent, and analytics run against live data instead of a nightly copy.
The limits are worth stating. TiDB Cloud Starter is multi-tenant, so the cheapest tiers are not isolated machines. Several of the newer plans, including PostgreSQL-compatible Starter, Essential, Premium, and BYOC, sit in public preview. Teams with a strict compliance footprint usually end up on Dedicated or BYOC rather than the entry tiers.
Getting Started
- Sign up for a TiDB Cloud account. No credit card is needed to start on Starter.
- Create a cluster. Pick Starter for a free serverless instance or Dedicated for a dedicated machine, then choose a region on AWS, Google Cloud, Azure, or Alibaba Cloud.
- Connect your app. Copy the connection string and use any MySQL-compatible driver or ORM. Migrating from an existing MySQL instance usually means changing the host, not the code.
- Load your data. Import a dump through the console, or point your existing MySQL replication at the new cluster during a migration.
- Monitor and scale. Watch query performance in the built-in dashboard, set a spending limit, and adjust compute or storage as traffic changes.
Product Information
A quick look at TiDB Cloud's pricing, supported platforms, and performance.
Best for
The users, tasks, and scenarios where this tool fits best.
Users
- Backend developers whose app is outgrowing a single MySQL instance
- Data teams that want to skip a separate analytics warehouse
- Teams building AI features on top of app data
- Small teams that want a real distributed database without a budget
Tasks
- Scaling writes past one server
- Running analytics on live data
- Migrating off manual sharding or a standalone warehouse
- Building semantic search into an app
- Handling spiky traffic
Scenarios
- A startup whose MySQL database is hitting connection and write limits during a growth spurt
- A team that built a dashboard on a nightly export because the production database couldn't take the queries
- An app adding a "search by meaning" feature
- A company that needs data to stay in its own AWS account
Key features
MySQL Compatibility Without a Rewrite
TiDB Cloud speaks the MySQL wire protocol, so existing drivers, ORMs, and SQL tooling connect as they are. For most teams that means migrating from MySQL is a connection change, not a code project. PingCAP also offers a PostgreSQL-compatible Starter instance in public preview, which matters if your stack leans Postgres rather than MySQL.
Distributed Scale With No Sharding
Data is split across storage nodes automatically, and your application never manages shard keys or routing. Scale-out happens by adding nodes rather than by rewriting queries. That's the core difference from sharding MySQL by hand: consistency and ACID transactions hold across the whole cluster, not just within one shard.
HTAP: Transactions and Analytics Together
TiDB Cloud is built for HTAP, combining transactional and analytical processing on one platform. A built-in columnar store answers analytical queries on current data, so you don't need a separate warehouse or an ETL pipeline feeding it. Analysts query the same cluster the app writes to, without slowing the app down.
Native Vector Search for AI Apps
Vector indexing and approximate nearest neighbor search are part of the database, not an add-on. Embeddings and relational records sit in the same cluster, and you can combine similarity search with SQL joins and metadata filters in a single query. For RAG pipelines, that means the retrieval layer is the database you already run.
Elastic Serverless on Starter
TiDB Cloud Starter is a serverless database that scales compute up and down with demand and down to zero when idle, so a prototype doesn't pay for peak capacity it isn't using. Newer Starter accounts offer 25 GiB of row storage, 25 GiB of columnar storage, and 250 million Request Units per month across eligible instances, with consumption billing beyond the free quota.
Deploy Anywhere Across Clouds
Clusters run on AWS, Google Cloud, Microsoft Azure, and Alibaba Cloud, across more than 30 regions, so you can keep data close to users and avoid a single vendor's lock-in. Enterprise-grade features include encryption in transit and at rest, private connectivity, and automated failover for business continuity.
Pros and cons
Pros
- MySQL compatibility means most migrations are a connection change rather than a rewrite, which keeps the switching cost low.
- Transactions and analytics run on the same cluster, so there's no separate warehouse to build or keep in sync.
- Vector search is built in, so AI apps don't have to run and reconcile a second database for embeddings.
- Serverless Starter scales to zero, which keeps idle and prototype costs down and needs no capacity planning.
- It runs on AWS, Azure, Google Cloud, and Alibaba Cloud, so the deployment isn't tied to one provider.
Cons
- The entry tiers are multi-tenant, so a team that needs hard isolation from other customers has to move up to Dedicated or BYOC.
- Several plans, including PostgreSQL-compatible Starter, Essential, Premium, and BYOC, are still in public preview, so some features can shift before general availability.
- Compared with a single small MySQL box, running a distributed database takes more moving parts to understand, which is real overhead for a small team.
- Published prices are usage-based and vary by plan and region, so estimating a monthly bill up front takes some work.
Frequently asked questions
TiDB Cloud runs applications that have outgrown a single MySQL server. It handles transactional writes and analytical queries on one distributed cluster, and it adds vector search for AI features, all as a managed database service.
Related content
Explore related tools, skills, and articles for TiDB Cloud.
TiDB Cloud Alternatives
Odyssey 3
Odyssey · OtherOdyssey-3 is Odyssey's foundation world model. It simulates physics in real time, turns a prompt into an interactive environment, and adapts to robots, cars and games.
JarvisCore
Prescott Data · Coding · Productivity · OtherJarvisCore is an open source Python runtime where AI agents work as equal peers on an agent mesh, claim tasks from a shared ledger, and reach external services through a zero-trust broker.
Clef
Cloudflare · Business · OtherClef is Cloudflare's open-source decision model for Workers AI. It returns calibrated probabilities for typed questions, not generated text, so agents can classify, route, and act fast.